Journal of Neurophysiology

Extended Voluntary Running Inhibits Exercise-Induced Adult Hippocampal Progenitor Proliferation in the Spontaneously Hypertensive Rat

Andrew S. Naylor, Anders I. Persson, Peter S. Eriksson, Ingibjörg H. Jonsdottir, Thorleif Thorlin

Abstract

Previous work has shown that voluntary running increases cell proliferation and neurogenesis in the dentate gyrus of the adult hippocampus. Here we report that long-term running for 24 days results in a down-regulation of hippocampal progenitor proliferation to one-half the level of nonrunning controls compared with a fivefold increase in progenitor proliferation seen after 9 days of voluntary running (short-term running). The negative effects seen on proliferation after 24 days of running were prevented by restricting daily running distances (by 30–50%) during 24 days. Long-term running for 24 days increases the response of the hypothalamic-pituitary-adrenal axis, with an increase in adrenal gland weight and increased plasma corticosterone levels, as well as decreased thymus weight, indicating a stress response as a possible mediator of decreased progenitor proliferation. Furthermore, the negative effects seen on the observed stress response after 24 days of running were prevented by restricting daily running distance. Short-term running did not alter these stress parameters compared with nonrunning controls. However, it increased phosphorylated cyclic AMP response element binding protein (pCREB) in the dentate gyrus, an increase that was not seen in nonrunning controls or after 24 days of running. Taken together, these data suggest that voluntary running does not always enhance proliferation and that the decrease in progenitor proliferation seen in long-term running is possibly mediated by mechanisms involving a stress response in the animal. However, a moderate level of long-term running was able to prevent the negative stress-related changes seen in unrestricted long-term running.

INTRODUCTION

New neurons are continuously added to the dentate gyrus of the hippocampus in adult mammalians such as rodents and humans involving the proliferation of neural progenitors in a process called neurogenesis (Altman and Das 1965; Eriksson et al. 1998; Gage et al. 1998). The hippocampus is a part of the brain primarily involved in memory and learning (Gould et al. 1999; Shapiro 2001), and the hippocampus has been identified as being vulnerable to stress and increased glucocorticoid levels (Cameron and Gould 1994; Kim and Diamond 2002). Voluntary exercise has been shown to increase the number of dividing cells in the granule cell layer (GCL) of the dentate gyrus in mice (Holmes et al. 2004; van Praag et al. 1999b), in treadmill-exercised rats (Carro et al. 2001), and in voluntary running rats (Persson et al. 2004). Voluntary exercise has also been shown to increase long-term potentiation and spatial memory in mice (van Praag et al. 1999a).

The mechanisms surrounding proliferation have focused on various growth factors important for progenitor proliferation in the hippocampus of running animals (Carro et al. 2001; Fabel et al. 2003). However, there has been particular attention devoted to the activation of the cyclic AMP response element binding protein (CREB) pathway, which is under the control of brain-derived neurotrophic factor (BDNF), in increasing adult hippocampal proliferation and neurogenesis (Nakagawa et al. 2002b), implicating a role for CREB in hippocampal neurogenesis during exercise.

The last decade has seen a wealth of papers studying the effects of adrenal steroids on hippocampal structure and function (Cameron and Gould 1994; Cameron et al. 1993; Gould et al. 2000; McEwen et al. 1992). It is known that adrenal steroids decrease hippocampal proliferation (Cameron and Gould 1994) and can have negative effects on hippocampal plasticity (Rodriguez et al. 1998). The loss of hippocampal neurons and hippocampal shrinkage related to elevated glucocorticoids has furthermore been linked with the pathogenesis of the depressive disease (Eriksson and Wallin 2004; Jacobs et al. 2000; Sheline et al. 1996). The hypothalamic-pituitary-adrenal (HPA) axis, the neuroendocrine system that is heavily involved in the body's response to stressful challenges by regulating glucocorticoid levels (Brown et al. 1999; Reul et al. 2000), has been shown to be modulated in voluntary running mice (Droste et al. 2003) and treadmill-exercising rats (Moraska et al. 2000).

We have previously shown that stress-related parameters are not changed after 9 days of voluntary running, and in these experiments, a positive effect on hippocampal progenitor proliferation is seen from running (Persson et al. 2004). However, long-term exercise has been related to stress perturbations, and this has led us to question if longer periods of running might at some stage become stressful and negatively impact on hippocampal progenitor proliferation. Therefore this study was undertaken to investigate whether a longer period of running activity would be stressful and affect the proliferation of newborn cells in the hippocampus of adult rats. As a possible mediator of the running induced proliferation, we also investigated changes in the levels of phosphorylated CREB (pCREB).

METHODS

Animals

The experiments were performed in female spontaneously hypertensive rats (SHR; Möllegaard Breeding Center), an in-bred rat strain chosen because of their tendency to run spontaneously in wheels. In terms of voluntary exercise, SHR run much further daily distances than their normotensive counterparts—the Wistar, Lewis, or Sprague-Dawley rat strains (unpublished observations). The rats arrived at the age of 13 wk and were kept for 7 days in the animal house for acclimatization before the onset of experiments. All animals were housed at a constant temperature (24°C) with relative humidity of 50–60%. A 12-h dark/light cycle was maintained with lights on from 7:00 a.m. to 7:00 p.m., with food and water available ad libitum. All procedures were approved by the Gothenburg committee for ethical review of animal experiments.

Voluntary running

Rats assigned to a running group had access to a running wheel (22.5 cm diam) attached to the side of the cage (41 × 31 × 23 cm). Wheel revolutions were automatically registered with customized computer software. The electronic locking equipment used in experiment 2 was a custom-built system attached to the side of each cage used for locking wheel access.

Bromodeoxyuridine labeling

To detect the proliferation of newborn cells, animals received intraperitoneal injections of bromodeoxyuridine (BrdU; 150 mg/kg) at 7:00 a.m. and 7:00 p.m. on the last 2 days of experiments. The last injection was given 12 h before death.

Experiment 1

The animals were randomized into four groups. Short-term runners (n = 12), consisting of 9 days of voluntary running; nonrunners (nonrunning controls for short-term runners; n = 6); long-term runners (n = 12), consisting of 24 days of voluntary running; and nonrunners (nonrunning controls for long-term runners; n = 6). All animals were weighed at the beginning and the end of their respective running periods.

Experiment 2

In an attempt to prevent the negative effects seen on hippocampal progenitor proliferation and stress responses in long-term runners, a running paradigm with restricted running activity over 24 days was used. Rats were divided into three groups: nonrunners (n = 8), long-term runners (n = 8), and locked-wheel runners (n = 8). Running rats were allowed to run for 24 days. The level of wheel activity in the locked-wheel group was determined by taking the average distance ran by short-term runners, i.e., 6 km/24 h. Therefore 2 h after the start of the dark period (9:00 p.m.), wheels were opened and remained open until the individual animal reached a level of 6 km, after which the wheel was automatically locked and stayed locked until 9:00 p.m. the next day. All animals were weighed at the beginning and end of the experiment.

Tissue preparation

Twelve hours after the last BrdU injection (7:00 a.m. on day 9 or day 24), the rats were anesthetized using isoflurane (Isoflurane Baxter, Baxter Medical, Kista, Sweden), weighed, and decapitated. After decapitation, the fresh brains were removed, and the left half of the brain was separated and fixed in 4% paraformaldehyde in 0.1 M phosphate buffer for 24 h and kept in a 30% sucrose solution at 4°C. Coronal sectioning (40 μm) was performed on a frozen microtome, and sections were stored in tissue cryoprotectant solution (25% ethylene glycol, 25% glycerol, and 0.1 M phosphate buffer) at –20°C to be used for immunohistochemistry analyses. Adrenal glands and thymus were removed and weighed. Trunk blood was collected immediately after decapitation within a 2-h period between 7:00 and 9:00 a.m., kept 30 min on ice (4°C), and centrifuged at 2,800g for 15 min. The plasma was stored at −20°C for further analysis.

Immunohistochemistry

The following antibodies and final dilutions were used to detect BrdU: primary monoclonal mouse anti-BrdU IgG antiserum (1:400; Boehringer-Mannheim, Germany), secondary horse anti-mouse IgG antiserum (1:125; Vector Laboratories, Kemila, Stockholm); to detect pCREB, primary rabbit anti-pCREB (1:600; Upstate Biotechnology), and secondary horse anti-rabbit IgG antiserum (1:125; Vector Laboratories, Kemila, Stockholm). Immunohistochemistry was performed on free-floating sections in Tris-buffered saline (TBS; 0.15 M NaCl and 0.1 M Tris-HCl, pH 7.5) pretreated with 0.6% H2O2 to remove endogenous peroxidase activity according to a previously published protocol (Persson et al. 2004). In brief, the sections were denatured in 50% formamide and 2× saline sodium citrate (SSC; 1× SSC, 0.3 M NaCl, and 0.03 M sodium citrate) for 2 h at 65°C, rinsed in SSC, followed by incubation in 2 M HCl at 37°C for 30 min, incubated 10 min in 0.1 M boric acid at pH 8.5, and followed by several washes with TBS. Before incubation overnight with the primary antibody, sections were blocked in TBS containing 0.25% Triton X-100 and 3% normal horse serum. Sections were washed in TBS, incubated with the secondary antibody, and allowed to react with avidin-peroxidase (ABC kit, Vectastain Elite, Vector Laboratories, Burlingame, CA) for 1 h. Finally, cells incorporated with BrdU or phosphorylated CREB were visualized using a detection solution (0.25 mg/ml diaminobenzidine, 0.01% H2O2, 0.04% NiCl) for ∼10 min. Differential interference contrast (DIC) images were obtained with a Nikon Diaphot Eclipse E600 microscope equipped with a CCD camera (Hamamatsu).

Quantification

Quantification of the number of BrdU-labeled or phosphorylated CREB-labeled cells in the GCL of the dentate gyrus were essentially performed as previously described (Aberg et al. 2000). For each animal, the number of positive cells in the GCL [including the subgranular zone (SGZ)] and their corresponding sample volumes were determined in 10 immunoperoxidase-stained coronal sections sampled throughout the whole length of the dentate gyrus. Images were taken with a Nikon Optiphot microscope connected to a video camera and area measurements of the GCL were performed with an Intuos Graphics tablet (Wacom) and digital image processing software (Nikon, Göteborg, Sweden). The number of positive cells were counted within the GCL and two cell diameters below the GCL, ignoring the cells in the uppermost focal plane and focusing through the thickness of the section to avoid errors caused by over sampling (Coggeshall and Lekan 1996; Gundersen et al. 1988). The number of positive cells were counted in each section and divided by the corresponding cross-sectional GCL area (given in μm) multiplied by the thickness of the section (40 μm) [BrdU/(area × thickness)]. The average number of positive cells per section was calculated for each rat and expressed as number of BrdU-positive cells or number of phosphorylated CREB-positive cells per cubic millimeter. To counter experimenter bias, slices were mounted and slides were coded, ensuring the experimenter was blind to which group the animal belonged to, and this code was not broken until all values were calculated.

Corticosterone measurement

The running rats have free wheel access and we know that their period of highest activity is during the very late period of the light phase and extends several hours into the dark phase. The acute effects of running on corticosterone were avoided by collection of blood 4–5 h after the last period of activity in these animals. Plasma samples from trunk blood were used for measurement of rat plasma corticosterone levels. A radioimmunoassay (RIA) based on 125I-labeled rat corticosterone was performed as previously described (Persson et al. 2004).

Statistical analysis

Values are expressed as mean ± SE. A plot of residuals against predicted values showed heteroscedasticity for the BrdU data, and logarithmic transformations were performed on these data. All other variables showed homoscedasticity, and log transformations were not needed. After these corrections, all data in experiment 1 were analyzed using a two-way ANOVA followed by posthoc comparisons using Bonferonni-Holm. In experiment 2, a one-way ANOVA followed by posthoc comparisons using a Tukey test was used. P < 0.05 was considered statistically significant.

RESULTS

Voluntary running can both increase and decrease hippocampal proliferation

Rats were individually housed and allowed free access to a running wheel. Short-term runners increased their running distance to 12.3 ± 2.3 (SE) km by the last day of running and, long-term runners increased and stabilized their running distance after 2 wk to 22.8 ± 1.9 km, which was maintained until the last day (Fig. 1). After 9 days of running, body weights did not differ between nonrunners (196.6 ± 2.6 g) and short-term runners (192 ± 3.1 g). Likewise, after 24 days of running, there was no difference in body weight between nonrunners (204.1 ± 3.8 g) and long-term runners (199.8 ± 3.7 g).

FIG. 1.

An illustration of 9 (SR) and 24 days (LR) of voluntary running activity in experiment 1. Data are expressed as daily mean running distance (km) per 24 h ± SE.

There was an increase in BrdU labeling from 7,354 ± 681 BrdU-positive cells/mm3 in nonrunners to 29,030 ± 2,096 BrdU-positive cells/mm3 in short-term runners (394 ± 18% increase compared with nonrunners, P < 0.001, Fig. 2A). In contrast, there was a decrease in BrdU labeling from 6,389 ± 526 BrdU-positive cells/mm3 in nonrunners to 3,110 ± 511 BrdU-positive cells/mm3 in long-term runners (48 ± 4% decrease compared with nonrunners, P < 0.05, Fig. 2A). In all groups, BrdU-labeled cells were distributed throughout the SGZ, with larger clusters of cells seen in short-term runners. After long-term running, very few BrdU-positive cells, usually seen as single cells, were observed in the SGZ (Fig. 2B). The average volume size of the GCL for nonrunning rats (0.79 ± 18 mm3) compared with short-term running rats (0.76 ± 20 mm3) did not significantly differ. There was also no difference between nonrunning rats (0.82 ± 21 mm3) and long-term running rats (0.82 ± 14 mm3).

FIG. 2.

A: nine days of voluntary running (SR) was a strong inducer of bromodeoxyuridine (Brd)U-positive cells in the granule cell layer (GCL) of the dentate gyrus compared with nonrunners (NR). However, 24 days of voluntary running (LR) greatly decreased BrdU-positive cells compared with nonrunners. B: cells labeled by BrdU were sparsely distributed throughout the GCL in nonrunners. After 9 days of running, the cells were widely distributed, usually observed in small clusters. After 24 days of running, few BrdU-labeled cells were observed in the GCL, usually as single cells. Scale bar, 100 μm. *P < 0.05, ***P < 0.001.

Role for stress-related parameters in running induced hippocampal proliferation

Short-term running did not affect adrenal gland weights compared with nonrunners. In contrast, long-term runners increased adrenal gland weights from 14.4 ± 0.6 mg/100 g body weight in nonrunners to 22.2 ± 0.9 mg/100 g body weight in long-term runners (P < 0.001). Similarly, a significant difference was found between short-term runners (14.5 ± 0.8 mg/100 g body weight) versus long-term runners (22.2 ± 0.9 mg/100 g body weight; P < 0.001; Fig. 3A). Circulating plasma corticosterone levels did not differ between nonrunners and short-term runners. Long-term runners increased corticosterone levels from 416.1 ± 44.4 ng/ml in nonrunners to 619 ± 51 ng/ml in long-term runners (P < 0.05). Similarly, a significant difference was found between short-term runners (447 ± 14 ng/ml) versus long-term runners (619 ± 51 ng/ml; P < 0.01; Fig. 3B). Thymus weights did not differ between nonrunners and short-term runners. Long-term runners had decreased thymus weights from 88.8 ± 5.1 mg/100 g body weight in nonrunners to 65.5 ± 3 mg/100 g body weight in long-term runners (P < 0.01). Similarly, a significant difference was found between short-term runners (81 ± 1.6 mg/100 g body weight) versus long-term runners (65.5 ± 3 mg/100 g body weight; P < 0.05; Fig. 3C).

FIG. 3.

To investigate the effect of voluntary running on the hypothalamic-pituitary-adrenal (HPA) axis, we measured adrenal gland weight, corticosterone levels, and thymus weights. A: short-term running (SR) did not affect adrenal gland size compared with nonrunners (NR). However, long-term running (LR) robustly increased adrenal gland size compared with nonrunners and short-term runners. B: corticosterone levels after short-term running did not differ compared with nonrunners. However, long-term runners increased their corticosterone levels compared with nonrunners and short-term runners. C: thymus weights did not differ between short-term runners and nonrunners. However, a decrease was observed in long-term runners compared with both nonrunners and short-term runners. **P < 0.01; ***P < 0.001.

Regulation of hippocampal phosphorylated CREB in the dentate gyrus of voluntary running rats

To further investigate a possible mechanism behind running induced hippocampal proliferation, the level of pCREB was analyzed using immunohistochemistry. There was a strong increase of pCREB labeling from 1,539 ± 500 pCREB-positive cells/mm3 in nonrunners to 5,402 ± 840 CREB-positive cells/mm3 in short-term runners (P < 0.001). Similarly, a significant difference was also found between short-term runners (5,402 ± 840 CREB-positive cells/mm3) versus long-term runners (2,764 ± 526 pCREB-positive cells/mm3; P < 0.001; Fig. 4, A and B). No significant difference in pCREB could be detected between nonrunners and long-term runners.

FIG. 4.

A: short-term running (SR) was a strong inducer of cAMP-response-element binding (CREB) phosphorylation in the GCL of the dentate gyrus compared with nonrunners (NR) and long-term runners (LR). B: pCREB labeling was seen specifically in the subgranular zone (SGZ) of the dentate gyrus. Scale bar, 100 μm. **P < 0.01, ***P < 0.001.

Restricted running activity prevents decreased BrdU-labeling in long-term running rats

In an attempt to prevent the negative effects seen in long-term running on hippocampal progenitor proliferation and stress responses in experiment 1, we restricted running activity in experiment 2 using a locking system to allow only 6 km of free wheel running per 24 h for 24 days. Body weights did not differ between nonrunners (192.5 ± 3.9 g) and long-term runners (193.0 ± 4.4g). However, there was a significant increase in body weight seen in locked-wheel runners (215.0 ± 3.9 g) versus nonrunners (192.5 ± 3.9 g) and long-term runners (193.0 ± 4.4 g; P < 0.01), respectively. Rats in both running groups quickly found the running wheels and by day 4 the locked-wheel group had reached 6 km/24 h, and that distance was achieved each day by all rats until day 24. Long-term runners without the locking device followed a similar pattern as in experiment 1, and by 2 wk, had stabilized their running activity to 22.2 ± 1.1 km/24 h. There was a similar decrease in BrdU labeling seen in long-term runners from experiment 1, from 4,805 ± 426 BrdU-positive cells/mm3 in nonrunners to 2,403 ± 558 BrdU-positive cells/mm3 in long-term runners (50 ± 2% decrease compared with nonrunners, P < 0.05). However, this decrease was prevented in the locked-wheel rats, showing no significant difference between nonrunners (4,805 ± 426 BrdU-positive cells/mm3) versus locked-wheel runners (4,342 ± 440 BrdU-positive cells/mm3). The BrdU labeling was also significantly higher in locked-wheel runners (4,342 ± 440 BrdU-positive cells/mm3) versus long-term runners (2,403 ± 558 BrdU-positive cells/mm3; P < 0.05; Fig. 5). The average volume sizes of the GCL for nonrunning, long-term running, and restricted running rats were 0.83 ± 16, 0.85 ± 21, and 0.82 ± 10 mm3, respectively. Restricted running did not significantly alter the volume of the GCL.

FIG. 5.

Long term running (LR) produced a decrease in BrdU-positive cells in the GCL of the dentate gyrus compared with nonrunners (NR) and locked-wheel runners (LW). Restricted running activity prevented the decrease seen in long-term runners was not significantly different compared with nonrunners.*P < 0.05.

Restricted running activity normalizes the stress responses seen in long-term running

To determine if the normalized BrdU labeling after restricted running activity was due to an altered stress response, we measured the same stress parameters as in experiment 1. As seen in experiment 1, adrenal gland weight was increased in long-term runners from 14.5 ± 0.6 mg/100 g body weight in nonrunners to 21.5 ± 1 mg/100 g body weight in long-term runners (P < 0.001). Adrenal gland weight in the locked wheel-runners was not significantly different compared with nonrunners, but was significantly lower (16.9 ± 0.9 mg/100 g body weight) compared with long-term runners (21.5 ± 1 mg/100 g body weight; P < 0.01; Fig. 6A). Circulating plasma corticosterone levels were increased in long-term runners from 368 ± 16.2 ng/ml in nonrunners to 631 ± 56 ng/ml in long-term runners (P < 0.05). There was no difference seen in corticosterone levels between nonrunners and locked-wheel runners. Locked-wheel runners had significantly lower corticosterone levels (376 ± 73 ng/ml) compared with long-term runners (631 ± 56 ng/ml; P < 0.05; Fig. 6B). Thymus weights were decreased in long-term runners from 95.3 ± 6.3 mg/100 g body weight in nonrunners to 62 ± 2.3 mg/100 g in long-term runners (P < 0.001). A similar decrease was also observed in locked-wheel runners from 95.3 ± 6.3 mg/100 g body weight in nonrunners to 72.1 ± 3.2 mg/100 g body weight in locked-wheel runners (P < 0.01), and there was no difference seen in thymus weights between long-term runners or locked-wheel runners (Fig. 6C).

FIG. 6.

To investigate the effect of restricted running activity on the HPA axis, we measured adrenal gland weight, thymus weight, and corticosterone levels. A: long-term running (LR) robustly increased adrenal gland weight compared with nonrunners (NR) and locked-wheel runners (LW). Locked-wheel runners did not increase their adrenal gland weights compared with nonrunners. B: corticosterone levels in locked-wheel runners were not significantly different compared with nonrunners. Long-term running increased CORT levels compared with nonrunners and locked-wheel runners. C: however, thymus weights were decreased in both long-term runners and locked-wheel runners compared with nonrunners.*P < 0.05, **P < 0.01, ***P < 0.001.

DISCUSSION

This study was designed to investigate how different voluntary running paradigms affect the proliferation of progenitor cells in the hippocampus of adult female SHR. We have previously shown the inducing effect on proliferation in the adult dentate gyrus after 9 days of voluntary running (Persson et al. 2004), and others have shown positive effects on proliferation in voluntary running mice (van Praag et al. 1999b) and treadmill-exercising rats (Trejo et al. 2001). In contrast, in this study, we show that hippocampal proliferation in the adult dentate gyrus is significantly decreased after 24 days of voluntary running compared with nonrunners. We also show how reduced daily running duration alters the response of 24 days of running by preventing the decrease of new progenitor cells seen in unrestricted 24-day runners.

It has been well established that the HPA axis plays a crucial role in the regulation of hippocampal proliferation (Cameron and Gould 1994; Gould et al. 2000). Decreased levels of glucocorticoids, i.e., corticosterone in rats, increases progenitor proliferation in the hippocampus (Gould et al. 1992; Yehuda et al. 1989), and increased levels of corticosterone decreases progenitor proliferation in the adult hippocampus (Cameron and Gould 1994). The HPA axis serves as a dynamic regulatory system for the release of corticosterone from the adrenal glands. It has previously been shown that under prolonged periods of stress, increased HPA axis activity and elevated corticosterone levels are accompanied by increased hypertrophy of the adrenal gland (Gomez et al. 1996; Sardessai et al. 1993). In this study, short-term running does not affect the size of the adrenal glands compared with nonrunning controls. However, in the long-term runners, an increase of about 65% in adrenal gland size was seen compared with nonrunning rats, which is considered a substantial stress-related response (Gomez et al. 1996). The adrenal glands did not increase in size in restricted runners showing a similar size compared with nonrunning rats. Corticosterone measured at the end of the running period was elevated in long-term runners with no changes in the short-term runners compared with nonrunning controls. The corticosterone levels were similar in restricted runners compared with nonrunning controls. Although measurement of corticosterone at one time-point does not allow us to see if corticosterone levels are persistently high throughout the circadian rhythm, the increased size of the adrenal glands is indicative of a chronic stress response in long-term running animals. The increased corticosterone levels and importantly the increased size of the adrenal glands found after long-term running in this study indicates a stress response in these animals compared with nonrunning controls and short-term running rats. This response could therefore be seen as one of the plausible explanations of the decreased hippocampal progenitor proliferation observed in the long-term running rats.

Another parameter related to a stress response is the thymus. Thymic involution occurs when corticosterone levels are chronically elevated (Akana et al. 1985; Hori et al. 1993) and is mediated through apoptosis of immature lymphocytes (Cohen 1992). In this study, the increased adrenal gland weight seen in long-term runners was accompanied with a decreased thymus weight, giving further indications of a stress response with elevated corticosterone levels in the unrestricted long-term runners. In contrast, short-term running did not affect adrenal gland weight or thymus weight, and an increase in hippocampal progenitor proliferation was seen. The restriction of daily running activity in long-term runners prevented an increase in adrenal gland weight and showed no difference in hippocampal progenitor proliferation compared with nonrunners. However, thymus weight was decreased in the restricted runners, possibly suggesting a small stress response that might explain why progenitor proliferation in these running animals does not exceed nonrunning rat levels.

Glucocorticoids, in particular corticosterone, have been shown to be involved in hippocampal granule cell apoptosis (Cameron and Gould 1994; Gould and McEwen 1993), and we speculate that part of the effect seen on decreased proliferation in the long-term running rats may be through elevated corticosterone levels, impacting on cellular death in the hippocampus. Heightened levels of corticosterone have also been shown to be a potent inhibitor of new cell birth in the hippocampus (Gould et al. 1992) and increased levels of corticosterone in the long-term running rats may also have an impact on cellular birth in these experiments.

The female SHR was chosen primarily because of their tendency to run long distances spontaneously in wheels in comparison to other rat strains. In comparison, Sprague-Dawley rats run ∼3 km/day after 3 wk of wheel running, similar to Wistar rats, whereas male SHR have been shown to run ∼9 km/day (Jonsdottir and Hoffmann 2000; Tong et al. 2001; Yamamoto et al. 2003). In comparison, in female SHR, we see a running distance around 23 km/day after 3 wk. Almost no studies have looked at the effect of long distance running activity on proliferation. However, a recent article using mice selectively bred for increased voluntary running found that increased running augmented neurogenesis in the hippocampus but reached a plateau (Rhodes et al. 2003). The estrus cycle has been shown to influence running activity. Over the estrus cycle, wheel running has been shown to be maximum under proestrus and minimal at metestrus (Kent et al. 1991). However, in this study, the interindividual running activity showed little differences during the running period, indicating that there might not be a strong influence from the estrus cycle on running in these rats.

The stress response in the SHR should be considered when relating these results to studies comparing voluntary running in various other rat strains. In response to stress, several different inbred rat strains, including the SHR, have been shown to present a similar increase in corticosterone levels and adrenal gland size and down-regulation of glucocorticoid receptor expression in a restraint stress protocol (Gomez et al. 1996). In these studies, SHR and Lewis rats had lower release levels of adrenocorticotropin hormone from the pituitary gland compared with other rat strains, which might reflect a somewhat altered feedback system in these two strains. However, the increase seen in stress-related parameters in treadmill running Wistar and Sprague-Dawley rats (Concordet and Ferry 1993; Moraska et al. 2000) and in voluntary running mice (Droste et al. 2003) are similar to what is seen here, suggesting that the observed stress reactions in the SHR has a relevance compared with other strains.

The literature concerning exercise effects on hippocampal proliferation has focused on the effects of various growth factors important for proliferation in voluntary running. Insulin-like growth factor I is a growth factor involved in normal neurogenesis (Aberg et al. 2000) and is also known to be of importance in running induced changes in neurogenesis (Trejo et al. 2001) and also acts as a neuroprotectant in the running animal (Carro et al. 2001). Another growth factor closely linked to voluntary running induced changes in proliferation and neurogenesis is BDNF (Farmer et al. 2004; Gomez-Pinilla et al. 2002; Neeper et al. 1996) and also vascular endothelial growth factor (Fabel et al. 2003). It is clear that these growth factors have a large influence on the proliferation of new cells in the hippocampus of running animals.

In short-term running, there seems to be no association between progenitor proliferation and the stress parameters measured in this study. Short-term runners showed an increase in CREB phosphorylation in the SGZ of the hippocampus that was not observed in long-term runners. Phosphorylation of this transcription factor has been shown to regulate normal adult hippocampal neurogenesis (Nakagawa et al. 2002b). Furthermore, it has been specifically shown that pCREB labeled cells are co-labeled with immature neurons in the SGZ, and following differentiation into mature neurons, pCREB is no longer expressed (Nakagawa et al. 2002a). The relationship between pCREB activation and progenitor proliferation in the hippocampus is not fully understood, but it is speculated that pCREB induction in immature neurons influences neighboring progenitor cells to increase rates of proliferation (Nakagawa et al. 2002b). In addition, it has recently been suggested that glucocorticoids decrease pCREB levels in cultured hippocampal progenitors (Yu et al. 2004). In this study, short-term running was a strong inducer of both hippocampal proliferation and pCREB, whereas long-term running showed a decrease in hippocampal proliferation and no change in pCREB-positive cells compared with nonrunners. These findings seem to be in line with previously discussed work showing a close correlation between hippocampal proliferation and expression of pCREB in the SGZ. A part of the decrease in hippocampal progenitor proliferation seen in the long-term runners is therefore suggested to be mediated by a down-regulation of pCREB during long-term running compared with the pCREB levels in the short-term runners. However, levels of pCREB in long-term runners did not decrease below nonrunning levels. This indicates that pCREB regulation might be of minor importance in the down-regulation of progenitor proliferation below nonrunning control rat levels and that other mechanisms, i.e., glucocorticoid-induced apoptosis or the effect of several growth factors discussed above may have a larger impact on this reaction.

In conclusion, these data show a biphasic effect of hippocampal progenitor proliferation after voluntary running in the adult rat that seems dependent on the level of running activity. A moderate amount of running may be more beneficial in stimulating progenitor proliferation in the hippocampus. Long periods of voluntary running seen here is a stressful event decreasing progenitor proliferation possibly through a general increase in several stress parameters. These findings provide an insight into the importance of further understanding the various impacts of exercise and stress on brain function and the need to clearly define voluntary exercise paradigms.

GRANTS

This work was funded by grants from the Institute for Stress Medicine (ISM; project 02–03), the Swedish National Centre for Research in Sports (CIF; project 60–03), the Swedish Society for Medical Research, Edit Jacobsson fund, and Vetenskapsrådet (VR; project 12x-12535).

Footnotes

  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

REFERENCES

View Abstract